Glucose distinctly affects cytosolic and mitochondrial NADH and NADPH in rodent beta cells: implications for the stimulation of insulin secretion

Hajj Hassan, Yara [UCL] Vandrisse, Clément [UCL] Tariq, Mohammad [UCL] Jonas, Jean-Christophe [UCL]

Background and aims: The stimulation of insulin secretion by glucose (GSIS) depends on accelerated glycolysis and mitochondrial metabolism in beta cells: the autofluorescence of NADH+NADPH increases across the whole range of glucose concentration between 0.5 and 30 mmol/l, with a small increase in the cytosol followed by a large increase in mitochondria, and this effect precedes and correlates with the rise in ATP production, Ca2+ influx and GSIS. More recently, it was shown that cytosolic NADPH amplifies insulin secretion via SENP1-mediated deSUMOylation of proteins involved in exocytosis, and that the rise in NADPH occurs in the cytosol and mitochondria of beta cells. However, the source of cytosolic NADPH and its role in GSIS remain disputed. Here, we used the NADPH sensors iNaps and NADH/NAD+ sensor SoNar to characterise the concentration-dependent effects of glucose on cytosolic and mitochondrial NADH and NADPH in two models of rodent beta cells and compared them to the Ca2+ and insulin secretion response curves. Materials and methods: Experiments were carried out in islets from C57BL/6 mice expressing functional nicotinamide nucleotide transhydrogenase and in rat INS 832/13 insulin-secreting cells cultured in RPMI medium containing 10 mmol/l glucose. Cytosolic and mitochondrial (mt-) NADPH levels and NADH/NAD+ ratios were measured using iNap1, mt-iNap3, SoNar or mt-SoNar expressed by adenoviral infection under the control of the rat insulin promotor for islets or the CMV promoter for INS cells. iNapC and mt-iNapC were used to control for the impact of pH and FAD autofluorescence on the results. The dynamic changes in probe fluorescence (exc 405/485; em 530) upon stepwise increases in glucose concentration and eventual addition of the mitochondrial uncoupler FCCP were monitored every 30 s. The experiments ended by maximal oxidation of NADPH with diamide or NADH with KP372-1. Results: At 0.5 mmol/l glucose, NADH and NADPH normalized to diamide or KP372-1 was larger in mitochondria than the cytosol. In both mouse islets and INS cells, mitochondrial NADH and NADPH progressively increased between 0.5, 2, 5 and 10 mmol/l glucose, with no further increase above10 mmol/l glucose. As expected, their levels rapidly decreased upon addition of FCCP. In comparison, cytosolic NADPH maximally increased between 0.5 and 2 mmol/l glucose in INS cells and between 0.5 and 5 mmol/l glucose in islet beta cells, while cytosolic Ca2+ and insulin secretion were still at basal level in each model. In contrast, cytosolic NADH remained low between 0.5 and 10 mmol/l glucose but markedly increased in response to 20-30 mmol/l glucose. Conclusion: Glucose distinctly affected NADH and NADPH in the cytosol and mitochondria of rodent beta cells: while mitochondrial NADH and NADPH increased in parallel with mitochondrial metabolism and metabolic amplification of insulin secretion, cytosolic NADPH was maximal at glucose concentrations below the threshold for Ca2+ influx and GSIS, and cytosolic NADH only increased at a very high glucose concentration. Our results are compatible with a permissive but not regulatory role of cytosolic NADPH in GSIS, and question the role of mitochondrial metabolism in cytosolic NADPH production in beta cells. They also suggest a possible role of high cytosolic NADH in beta cell glucotoxicity. Supported by: FRS-FNRS, ARC/FWB Disclosure: Y. Hajj Hassan: None.